论文部分内容阅读
目的探讨ARIMA模型设计流程和调整优化问题,对部分甲乙类传染病发病率进行预测。方法根据国家公布的1990-2000年的痢疾、伤寒、布鲁菌病和流行乙脑时序数据,作纯随机性和平稳性检验,低阶差分平稳序列绘制自相关和偏自相关图,借助SAS软件实现模型定阶调试、参数识别、残差检验、模型拟合预测。结果发病率数据存在短期相关性,一阶差分提取趋势信息,经BIC准则定阶疏系数模型,逐步经验调试与优化,参数显著性和残差白噪声检验有效,模型预测有实用价值。结论 ARIMA法对于传染病发病率等卫生预测问题有适用性、代表性,建模设计应针对实际资料进行综合研究。
Objective To explore the ARIMA model design flow and adjustment and optimization problems, and predict the incidence of some Class A and B infectious diseases. Methods According to the published data of dysentery, typhoid fever, brucellosis and epidemic encephalitis from 1990 to 2000 in our country, we used pure randomness and stationary test, and autocorrelated and partial autocorrelation diagrams of low-order differential stationary series. Software Implementation Model Ordering, Parameter Identification, Residual Testing, Model Fitting Prediction Results There was a short-term correlation between the incidence data and the first-order difference extraction trend information. The BIC criterion was used to determine the order and sparseness coefficient model, and the empirical test and optimization were carried out gradually. The parameter significance and the residual white noise test were effective. Conclusion The ARIMA method has applicability and representativeness to the health prediction problems such as the incidence of infectious diseases. The modeling and design should be comprehensively studied according to the actual data.